Reply to: Association Between Alendronate and All-Cause Mortality and Cardiovascular Mortality Among Hip Fracture: An Alternative Explanation.

To the Editor: We are thankful to Prof. Nguyen and Dr. Tran for their interest in our study and we appreciate the opportunity to respond to their comments. As Prof. Nguyen and Dr. Tran suggested that censoring patients at the time of switching medications might have inflated the effect size, we investigated this potential bias by excluding patients with switching of medications. Of the 3,081 alendronate-treated patients, 281 patients (9.1%) switched the therapy during the study period. After excluding these patients, we observed similar findings (Table), suggesting that bias due to treatment of censoring data should be minimal in our study.